This project precipitated from the problem of trying to identify a good used car on the open market. There are a number of factors that make this a challenge. Often times, vehicles are listed for more than their market value and are promoted heavily. This causes good ads to be overlooked. Curbsiders, as they are named, also tend to buy used cars cheaply from semi-private markets. These vehicles are usually poor value for the price that they command. Also, cars that have been substantially abused and modified make their way into the market. It is not always clear how to identify a good used car ad. This project aims to help out a prospective buyer.
The primary goals of this project are to learn how to:
- Scrape Canadian used car market data from ads in the following locations
- Kijiji
- Craigslist (possibly)
- Use build automation software
- Handle collaborative software development
- Communication
- Delegation
- Documentation
- Use software architecture patterns
Possible secondary goals:
- Market analysis with statistical methods
- Use NLP to predict the quality of an ad on the following points
- Price is below market mean
- Vehicle condition is acceptable
This project needs help with the following:
- OO design
- Flexible Swing GUI
- Testing
- Documentation
How to help
- Get in touch with me at k47swp at mun.ca